Financial inclusion workshop

IGC Mozambique hosted a workshop on evidence-based policymaking for financial development. The workshop was jointly held with the Bank of Mozambique, NOVAFRICA, NOVA School of Business and Economics , and GLM LIC.

This conference included a presentation on the research results and policy recommendations derived from the study “Determinants of Adoption of Electronic Means of Payment”, which is the result of a collaboration between the IGC and the Bank of Mozambique. The conference also featured discussions with a wide range of presenters and panellists, including Professor Dean Yang (University of Michigan and NOVAFRICA), Professor Dean Karlan (Yale University and NOVAFRICA), Dr Pedro Vicente (Nova School of Business and Economics, NOVAFRICA, and IGC), Catia Batista (Nova School of Business and Economics and NOVAFRICA), and Dr Sandra Sequeira (LSE and IGC).

Dr Esselina Macome (Bank of Mozambique) opened the workshop by emphasising the importance of the collaboration between the IGC and the Bank of Mozambique when it comes to understanding the adoption, use and impact of mobile money in Mozambique. Dr Macome outlined the development of electronic payment systems and regulation in Mozambique, which has happened alongside the expansion of the cell phone network. In the last few years, there has been an increase in financial services that can be accessed by cell phone. This has strongly contributed to financial inclusion by providing easy access to services and products for a substantial part of the adult population, especially in areas where it is difficult to access traditional financial services. In order to further continue the work on financial inclusion, collaboration with a range of actors will be crucial.

Dr Claudio Frischtak (IGC Mozambique) echoed Dr Macome’s appreciation of the successful collaboration between IGC and the Bank of Mozambique over the last 4-5 years in the area of financial inclusion. He emphasised the importance of coming together to discuss the results and to identify the way forward for ensuring that financial inclusion is effective in reducing poverty.

Ms Artemisia Gove (Bank of Mozambique) presented the results of a study on the determinants of the adoption of electronic payments in Mozambique, which was performed by the Bank of Mozambique together with NOVAFRICA, under funding from the IGC. The study focused on access to financial services for SMEs in the urban areas of Maputo and Matola, which is considered a key requirement for these companies to expand. The use of electronic payments is still limited. Out of the 1027 firms covered, only 22% used point of sales (POS) services, 4% used mobile banking, and 7% used mobile money. The main determinants for the use of POS were firm characteristics such as size, type, stability, security and access to electricity; the acceptance of electronic payments through mobile phones (mobile banking and mobile money) depends more on characteristics of the business owner, such as age, education and place of birth.

Across the various means of electronic payments, the study highlights the importance of distinguishing between access and use of financial services. Access appears not to be the main constraint to adoption, as the vast majority of the surveyed business owners (75%) have a bank account. The most frequently cited reason for not adopting electronic means of payments is a lack of knowledge about these types of services and how to access them. For example, 26.3% do not know what POS is or how it is used; for mobile banking, this figure was 35.98%.

Among those who have adopted electronic means of payment, satisfaction with the services is generally high. For example, almost half of the respondents using mobile money referred they had never had any problems with the service. However, there are also significant problems reported, including excessive bureaucracy, high fees and the quality and maintenance of networks.

To further spread the adoption of electronic means of payments, it is important to emphasise the advantages of using these services from a business perspective, and to promote and explain the correct use of the services. Ms Gove also recommended government intervention to improve coordination among all actors involved, as well as to reduce the time required by commercial banks to adhere to electronic systems (for POS).

In the discussion that followed the presentation, Dr Macome mentioned that the Bank would like to roll out this study in other cities, and may try to consult other actors about the questions they may want answered, as there was a great interest and requests for further details from a range of actors in the workshop, including mobile money operators and banks.

Sandra Sequeira (LSE and IGC Mozambique) then continued with a presentation on “The Role of Technology and Business Skills in Microenterprise Development: the Case of Mozambique” (ongoing study with Cátia Batista and Pedro Vicente). Over 50% of the urban poor are engaged in some form of microbusiness, and recent experience with mobile banking suggests that low-cost technologies such as cell phones could be an effective means to increase access to financial products and enhance the savings capacities of the poor, which can play a critical role in overcoming credit constraints to the growth and survival of their microbusinesses. However, experiences with such technologies have also highlighted the importance of having the business skills to effectively mobilize and apply these savings for business growth.

The study, covering 1,200 micro entrepreneurs in Maputo and Matola, seeks to answer two key research questions: 1) What is the impact of access to interest-accruing savings accounts through mobile money; and 2) What is the role of business management skills in microenterprise development and growth? Finally, the study seeks to identify the combined effects of savings products and business management skills.

The impact of three different types of intervention is tested; MKesh intervention (adoption of mobile savings account and initial incentives for utilising the service and for saving), business training intervention, and a combination of the two. Although the study is not yet concluded, the preliminary results are encouraging. The business training was well received, and increased the micro-entrepreneurs’ understanding of basic accounting. Furthermore, it reduced transfers to relatives by 14% and increased total sales by 64%. The strongest effect came from the combined treatment of mKesh adoption/use as well as business training. The impact on saving and expenditure is still unclear, and the final results may show more details on this, as well as the mechanism through which increased sales are occurring.

Dean Yang (University of Michigan) presented a study on “Smart Subsidies: How Combining Subsidies with Savings Brings Expanded Benefits to Rural Households” (joint study with Michael Carter and Rachid Laajaj).

A number of anti-poverty programmes are ‘bundled’; they consist of multiple components that are thought to complement each other, with the impact of the joint interventions being greater than the sum of impacts when offered separately. This study tries to understand the interplay of two interventions, agricultural input subsidies and formal savings, in Manica province. There were three savings treatment groups; basic savings, matched savings, as well as the control group. Within each of these groups, a subset of farmers was chosen randomly to also receive a voucher subsidising the purchase of a set of agricultural inputs.

Looking at the results, each type of treatment was successful at raising fertiliser use (although the impact was small and statistically insignificant for the basic savings treatment group alone). Similarly, all treatments increased savings, most dramatically in the treatment groups that received both the voucher as well as one of the savings interventions.

When looking at the impact of the interventions on household wellbeing (measured by daily consumption by capita), however, the results are different; in this case, the subsidies and savings programmes appear to be substitutes rather than complements, with the impact of the joint treatment being lower than the sum of the separate impacts. The results may be explained by understanding the dual role that savings can play in terms of investment and risk management.  Households receiving only a savings treatment use their savings for both purposes, while households receiving both subsidy and savings treatments seem to focus on risk management, leading to no additional impact in terms of the level of consumption.

The results also suggest that savings help households cope with negative agricultural shocks, as household consumption was more sensitive to these types of shocks when receiving the subsidy only. The consumption variance is lower in the savings treatment groups.

Pedro Vicente (NOVAFRICA, Nova University of Lisbon and IGC) presented the preliminary results of a study on “Agricultural Savings and Network Pressure: Experimental Evidence Using Mobile Money in Mozambique” (joint with Cátia Batista and Dean Yang). Savings are crucial to break the cycle of low investment and low agricultural productivity for farmers. The mobile money revolution underway in Africa seems like an opportunity to extend access to savings instruments. This study explores whether access to custom-made savings accounts offered through mobile money increases savings by farmers, and also looks at the effects on fertiliser use. Finally, the study examines whether access to a savings account can counter social pressures to share resources by shielding farmers from such pressures.

The study was carried out with 196 maize farmers in the province of Manica. All of them were introduced to MKesh and also got a module on fertilizer use. The two treatments were the savings treatment and the network treatment, which interacted in a 2 x 2 research design. The savings treatment offered a bonus of 20% interest on the average mobile balance held over a certain period. The network treatment offered the modules on mobile money and fertilizer use to two of the farmer´s closest friends.

The preliminary results show clear effects of access to the savings account on a number of fronts: 1) increasing mobile money adoption; 2) increasing non-frequent expenditures whilst decreasing the likelihood that individuals lend money to their closest farming friends; and 3) increasing fertiliser adoption. Finally, the data suggests that having access to a savings account counteracts social pressure to share resources.

Furthermore, the study indicates that incentivised mobile money agents are vital for both communication and adoption of mobile money; the mere existence of the service is not enough. It also suggests that more work can be done to extend access to interest-bearing accounts using mobile money platforms.

Cátia Batista (NOVAFRICA, Nova University of Lisbon) presented on “An Experimental Impact Evaluation of Introducing Mobile Money in Rural Mozambique” (joint study with Pedro Vicente). Access to financial services in rural areas of Mozambique is still very limited, and mobile money can be seen as a means of financial inclusion in these areas. The main research question in the study is the economic impact of introducing access to mobile money to people in rural areas in Southern Mozambique. The study involved a randomized field experiment, involving 102 locations, with mobile money being introduced in 51 of them using newly recruited mobile money agents and community-wide and individual dissemination.

The preliminary results show good levels of adoption of mobile money; 63% of individuals in treatment areas performed at least one mobile money transaction in the first year after the intervention. Although the number did decrease over the next two years, it did not fall dramatically. Over the period studied, there was also an evolution visible in terms of the types of transactions performed, with the proportion of transactions of airtime purchases decreasing, and an increase in transactions related to remote payments (e.g. electricity) and transfers being received. When it comes to the average number of mobile money transactions, there were no clear trends in the data, though peaks were seen which correspond to the timing of the maintenance of agent networks, highlighting the importance of a well-functioning network and customer support. Similarly, there were no obvious patterns in relation to the average value of mobile money transactions, though there seem to be peaks in the lean season, consistently with the literature on mobile money as an instrument for risk sharing.

In terms of transfers, the probability of the treatment group receiving transfers was higher. The data showed no results in terms of affecting the overall amount of savings, and aggregate consumption does not change significantly. At the same time, individuals in the treatment groups reported being less vulnerable to lack of access to water and lack of medical care, pointing to the role of mobile money in diminishing vulnerability to shocks. In light of the data, this reduced vulnerability is likely to happen through received remittances as savings are not significantly affected

Overall, challenges remain when it comes to the effective utilisation of mobile money, which requires investment in a well-functioning agent network and customer support.

The first discussant, Aurélio Matavel (mKesh), highlighted that access is still a very important element when it comes to populations living in rural areas, relating to the major challenge of establishing and servicing agent networks. Another key challenge is establishing and increasing trust in a system of electronic payments. Moving from cash to electronic means of payments is a change of paradigm, and so the process will take time, even in urban areas.

The second discussant, Dylan Lennox (M-Pesa), sees mobile money and mobile banking converging in a number of ways recently. He also emphasised that whilst competition in the market is good, collaborative competition in the area of financial inclusion is both critical and good due to the high costs involved of building agent networks and customer education. As such, collaboration across different actors and institutions is welcomed. Mr Lennox also summarised some of the key lessons from the earlier presentations, and emphasised in particular that the technology is necessary but not sufficient. Having a solid and sustainable network and infrastructure is crucial, as well as continued servicing of agents and education of both agents and customers.

Dean Karlan (Yale University and Innovations for Poverty Action) delivered a presentation entitled “Financial Inclusion: A Means to an End”. Professor Karlan emphasised the importance of remembering that financial inclusion in itself is not an end; statistics often cited do not actually tell us about the quality of the services or whether it satisfies the needs of the population. It is often difficult to know what to do or what outcomes to measure. Furthermore, many studies in the past show correlation, and not causation. To highlight this, Professor Karlan used evidence from seven RCTs in seven different countries exploring the impact of the traditional microcredit model. The model, whilst it may have other positive effects, does not effectively help with poverty alleviation. Professor Karlan emphasised the importance of two guiding principles for designing programmes: 1) evidence trumps hope and intuition; and 2) it is essential to design for humans and how they actually behave, rather than how one might think they should behave.

Some alternative microcredit programs with certain adapted loan features that are currently being implemented show encouraging signs in terms of impact, highlighting the importance of making the lending model fit better to the actual needs people have. Similarly, interventions in the areas of micro-savings and micro-insurance that incorporate features which address both transaction costs and behavioural factors are also showing encouraging results so far.

Pedro Vicente (IGC, NOVAFRICA, and Nova University of Lisbon) concluded the workshop, thanking the Bank of Mozambique for the successful collaboration and the opportunity to contribute to knowledge on the Mozambican economy. Aurora Bila (Bank of Mozambique) closed the conference, noting that the presentations generated a lot of interest across the spectrum of actors and that the Bank of Mozambique would like to continue the partnership to follow-up on already completed studies as well as undertake research in new areas of interest.

Overall, the studies presented at the workshop generated a lot of meaningful discussions on the respective roles of the various actors and stakeholders in the mobile money environment. The debate left no doubt that the perceptions on mobile money have evolved in recent years. Whereas the banking sector may at first have been hesitant about the prospect of mobile phone companies entering its traditional sphere of activities, the discussions during the workshop really recognized the complementarities between traditional banks and mobile money operators. At the same time, the banks are now also starting to develop new products resembling mobile money. As such, the relationship will surely remain a dynamic one, with complementarities across the board while competing in specific sub-areas. The IGC Mozambique has played a critical role in shaping these discussions through its 4 year research collaboration with the Bank of Mozambique. This workshop was a prime example of how the IGC can actively contribute to an effective and impactful dialogue among market players, with a view to strengthening financial inclusion in Mozambique.